Will the ratio of inference runs to training runs on GPT5 decrease from the ratio on GPT4? | Manifold

Will the ratio of inference runs to training runs on GPT5 decrease from the ratio on GPT4?

0

Ṁ100

Jan 1

50%

chance

1H

6H

1D

1W

1M

ALL

Market context

Get

1,000

to start trading!

People are also trading

Will the performance jump from GPT4->GPT5 be less than the one from GPT3->GPT4?

What will the aggregate improvement of GPT5 be over GPT4 in terms of metrics?

Will growth in the maximum MW used to train AIs, slow down by more than x2 after GPT-5-like?

Will GPT5 show clear signs of diminishing returns?

How much compute will be used to train GPT-5?

Will growth in the maximum MW used to train AIs, slow down by more than x2 after GPT-6-like?

GPT-5 plus scaffolding and inference-compute ~= training compute will achieve capabilities advance >= (GPT-4 to GPT-5).

Will it be possible to disentangle most of the features learned by a model comparable to GPT-4 this decade?

Related questions

Will the performance jump from GPT4->GPT5 be less than the one from GPT3->GPT4?

What will the aggregate improvement of GPT5 be over GPT4 in terms of metrics?

Will growth in the maximum MW used to train AIs, slow down by more than x2 after GPT-5-like?

Will GPT5 show clear signs of diminishing returns?

How much compute will be used to train GPT-5?

Will growth in the maximum MW used to train AIs, slow down by more than x2 after GPT-6-like?

GPT-5 plus scaffolding and inference-compute ~= training compute will achieve capabilities advance >= (GPT-4 to GPT-5).

Will it be possible to disentangle most of the features learned by a model comparable to GPT-4 this decade?

© Manifold Markets, Inc.•Terms•Privacy